Effective Positive Negative Association Rule Mining Using Improved Frequent Pattern

Association Rule is an important tool for today data mining technique. But this work only concern with positive rule generation till now. This paper gives study for generating negative and positive rule generation as demand of modern data mining techniques requirements. Here also gives detail of “A method for generating all positive and negative Association Rules” (PNAR). PNAR help to generates all unseen comparative association rules which are useful for interesting pattern finding. This work focus on determine positive and negative rules, generation of candidate set is key issue in these techniques. This paper also discussed existing techniques, such as frequent pattern growth (FP-growth) method it’s a most efficient and scalable approach for rules generation. This method can generate rules without candidate ser generation. This main problem in FP tree growth is large number of conditional FP tree. This algorithm able to generates all positive and negative association rule mining. We also proposed new positive and negative association rule mining algorithm using improved frequent pattern tree for better and efficient association rules.